How much fluid is down there
Gallons per foot of pipe is the bore squared times 0.0408. At a 1.076 inch bore that is 0.0472 gallons a foot, so 2,100 feet of pipe holds 99.2 gallons, and with three gallons in the heat pump and the flow centre the loop is 102.2 gallons. People underestimate this consistently, because a coil of pipe on a pallet does not look like a hundred gallons of anything.
Use the nominal size instead of the bore and the answer goes wrong fast, because the diameter is squared. Nominal one inch against an actual 1.076 inch bore is 16 percent, which on this loop is 14 gallons of fluid nobody ordered. On a job where the concentrate is the second largest line item after the drilling, that matters.
The blend has a closed form
Filling an empty loop is trivial: the product is the volume times the target divided by the product strength, and the rest is water. At 102.2 gallons, a 20 percent target and neat concentrate, that is 20.44 gallons of product and 81.76 of water.
Correcting a full loop is the interesting case, because there is nowhere to put anything until something comes out. Drain D gallons, put D gallons of neat product back, and the concentration goes from c0 to c0 plus D over V times the difference between the product and c0. Setting that equal to the target and solving gives D equals V times the target minus c0, all divided by the product strength minus c0. On a 102.2 gallon loop sitting at 8 percent, going to 20 percent with neat concentrate, that is 102.2 times 12 over 92, which is 13.33 gallons out and 13.33 gallons of product in.
That is the smallest drain that reaches the target using neat product for the entire refill. You can drain more and make the difference up with water; the product quantity is unchanged and the only thing you gain is a bigger bucket of waste fluid to dispose of.
Going the other way
Diluting is simpler and the page handles it. Draining D gallons out of a loop at c0 and replacing with water takes the concentration to c0 times one minus D over V, so getting from 25 percent down to 20 takes a drain of V times 5 over 25, which is a fifth of the loop. Nobody plans to over-concentrate a loop, but it happens, usually when a premixed fluid gets treated as concentrate on the fill.
The two numbers this page will not give you
It does not tell you what concentration to use and it does not tell you what temperature a given concentration protects to. Both come from the data sheet for the specific product, and both are more slippery than they look. Freeze point and burst point are different rows in the same table and they are typically a long way apart, because a fluid can turn slushy without splitting a pipe. Which of the two the designer specified changes the concentration substantially, and reading the wrong row is a real and common error.
Refractometers add their own trap. A refractometer has separate scales for different fluids and reading a propylene glycol sample on the ethylene glycol scale gives a confident, wrong answer. Check which scale is being read before believing what the loop is at, particularly on a loop somebody else filled and nobody documented.
The last thing worth naming is that what may go in the ground at a given site is not a technical question. Antifreeze in a buried loop sits inside the ground the local groundwater authority governs, and what is permitted there is decided by that authority. It varies, and it is not always the fluid that performs best or costs least.
Questions people ask
How many gallons of antifreeze does a ground loop need?
The loop volume times the target concentration divided by the strength of the product. At the defaults here — 2,100 feet of 1.076 inch bore plus three gallons of equipment, so 102.2 gallons, filled to 20 percent with neat concentrate — it is 20.44 gallons of product and 81.76 gallons of water. Change the bore, the run, the target or the product strength and it moves proportionally with all four. The target concentration comes from the fluid data sheet at the protection your designer specified, not from this page.
How do I raise the concentration in a loop that is already full?
Drain some out and put neat product back in its place. The smallest drain that reaches the target is the loop volume times the target minus the current concentration, divided by the product strength minus the current concentration. A 102 gallon loop reading 8 percent and heading for 20 with neat concentrate needs 13.3 gallons drained and 13.3 gallons of product back. Draining more than that works too, with water making up the balance, and uses exactly the same amount of product.
How much fluid does a foot of pipe hold?
The inside diameter in inches squared, times 0.0408, gives gallons per foot. A 1.076 inch bore holds 0.0472 gallons a foot; a 1.358 inch bore holds 0.0752. Use the bore from the pipe data sheet rather than the nominal size, because the figure is squared and the gap between nominal and actual is not small. Add whatever the heat pump, the flow centre and the hoses hold on top, from the equipment data.
What concentration do I need?
That is a question for the fluid data sheet and the person doing the design, and this page deliberately does not answer it. What the sheet gives you is a table of concentration against freeze point and against burst point, and those are different rows that are usually many degrees apart. Which one the design was based on, and what temperature the loop fluid is expected to reach at the worst point of the worst winter, are design outputs. The field on the form takes whatever you were given and does the blending arithmetic for it.
Why is my refractometer reading different from what I mixed?
Most often because it is being read on the wrong scale. A refractometer carries separate scales for different fluids, and a propylene glycol sample read against the ethylene glycol scale gives a plausible number that is simply not the concentration. The other usual reasons are that the loop was not circulated long enough to mix before sampling, that the sample came from a leg still holding fill water, or that the product was a premix being treated as concentrate. Check the scale first, then circulate and resample.